Exercise systolic blood pressure indexed to work rate predicts future risk of incident stroke

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): 1) The CircM network, Linköping University 2) The Swedish Stroke Association. Background Hypertension is a major risk factor for stroke. Exaggerated peak systolic blood pressure (SBP) during exercise is associated with increased risk of hypertension in normotensive subjects, whereas low, rather than high, peak exercise SBP predicts cardiovascular and all-cause mortality. Purpose We aimed to study the risk of incident fatal and non-fatal stroke in relation to absolute and work rate indexed exercise SBP, in patients referred for exercise testing. Methods We cross-linked clinical data from 14,428 bicycle ergometer exercise tests with data from the National Patient Registry and the Registry of Prescribed Medications up until December 2019. Subjects with a history of cerebrovascular disease were excluded, as well as subjects with submaximal tests, a drop in SBP during exercise, or with missing SBP data at the end of exercise. Baseline comorbidity was defined as diagnosis of, or treatment for, hypertension, diabetes, hyperlipidemia, and coronary artery disease (CAD). Work rate indexed SBP response (SBP/Watt-slope) was calculated as the increase in SBP during exercise divided by the increase in work rate. We applied Cox regression to calculate hazard ratios (HR) with 95% confidence intervals (95% CI) for the risk of incident stroke in association to peak SBP and the SBP/Watt-slope. We used absolute values as well the upper limit of normal (≥2 Z scores) of a recently published reference population, with reference equation including age, sex, SBP at rest, Wmax (for SBP peak only) and height (for SBP/W-slope only). Analyses were adjusted for age, sex, and baseline comorbidity. Results In total, 7,946 subjects (54% male, age 55±13 years) were included, with baseline comorbidity of hypertension (28%), diabetes (7%), hyperlipidemia (16%), and CAD (13%). During a median follow-up of 8.4 years (inter quartile range 5.8-11.5), 3.1% developed ischemic (2.2%) or hemorrhagic (1.0%) stroke. There were positive associations to future stroke with absolute peak exercise SBP (HR: 1.08 [1.03 – 1.13] per 10 mmHg increase) and with SBP/Watt-slope (HR: 1.09 [1.05 – 1.14] per 1 mmHg/10 Watt increase), but not after adjustment for sex, age and baseline comorbidity. However, having a SBP/Watt-slope ≥2 Z scores was associated with an increased risk for stroke (HR: 1.56 [1.08 – 2.25)] after adjustment for baseline comorbidity, whereas an adjusted SBP peak ≥2 Z scores was not (HR: 0.62 [0.32 – 1.21]). Conclusion A workload-indexed SBP response (SBP/Watt-slope) above the upper limit of normal may help to identify subjects with no history of cerebrovascular disease at risk of future stroke. In contrast, the absolute exercise SBP response to exercise does not add substantial information to the risk estimation above age, sex, and baseline comorbidity.